Fix #24 app.py, dependencies

app.py

@@ -11,7 +11,7 @@ from torchvision.utils import save_image
 import gradio as gr
 import numpy as np
 import io
-import tempfile  # Importar tempfile
+import tempfile
 
 # Asegúrate de que las funciones necesarias estén definidas (si no lo están ya)
 def resize(img, size):
@@ -266,13 +266,19 @@ class Solver(object):
                 transforms.Normalize(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5])
             ])
             # Convertir a PIL image antes de la transformación
-            with tempfile.NamedTemporaryFile(suffix=".jpg") as source_temp_file, \
-                    tempfile.NamedTemporaryFile(suffix=".jpg") as reference_temp_file:
-                source_temp_file.write(source_image)
-                reference_temp_file.write(reference_image)
-                source_image_pil = Image.open(source_temp_file.name)
-                reference_image_pil = Image.open(reference_temp_file.name)
-
+            try:
+                with tempfile.NamedTemporaryFile(suffix=".jpg") as source_temp_file, \
+                        tempfile.NamedTemporaryFile(suffix=".jpg") as reference_temp_file:
+                    source_temp_file.write(source_image)
+                    reference_temp_file.write(reference_image)
+                    # Verificar el contenido de los archivos temporales
+                    print(f"Primeros 100 bytes de {source_temp_file.name}: {open(source_temp_file.name, 'rb').read(100)}")
+                    print(f"Primeros 100 bytes de {reference_temp_file.name}: {open(reference_temp_file.name, 'rb').read(100)}")
+                    source_image_pil = Image.open(source_temp_file.name)
+                    reference_image_pil = Image.open(reference_temp_file.name)
+            except Exception as e:
+                print(f"Error al procesar las imágenes: {e}")
+                raise  # Re-raise la excepción para que Gradio la capture
             source_image = transform(source_image_pil).unsqueeze(0).to(self.device)
             reference_image = transform(reference_image_pil).unsqueeze(0).to(self.device)
 
@@ -312,7 +318,7 @@ def gradio_interface():
     )
 
     # Ruta al checkpoint
-    checkpoint_path = "iter/20500_nets_ema.ckpt"
+    checkpoint_path = "iter/10500_nets_ema.ckpt"
 
     # Crear la interfaz de Gradio
     inputs = [
@@ -335,4 +341,4 @@ def gradio_interface():
 
 if __name__ == '__main__':
     iface = gradio_interface()
-    iface.launch(share=True)
+    iface.launch(share=True)

requirements.txt

@@ -3,4 +3,5 @@ torch
 torchvision
 Pillow
 numpy
-huggingface_hub
+huggingface_hub
+tempfile